- Email: [email protected]
Sulfitolysis and thioredoxin-dependent reduction reveal the presence of a structural disulfide bridge in spinach chloroplast fructose-1,6-bisphosphatase
FEBS 19917
FEBS Letters 424 (1998) 109^112
Sul¢tolysis and thioredoxin-dependent reduction reveal the presence of a structural disul¢de bridge in spinach chloroplast fructose-1,6-bisphosphatase Dorothee F. Drescher, Hartmut Follmann, Ingo Haëberlein* University of Kassel, Department of Biochemistry, Heinrich-Plett-Str. 40, D-34109 Kassel, Germany Received 28 November 1997; revised version received 30 January 1998
Abstract A significant difference between cytosolic and chloroplastic fructose-1,6-bisphosphatase (FbPase) is an extra peptide in the middle of chloroplast FbPase which contains three additional cysteine residues. Sit-directed mutagenesis experiments have shown that at least two of these cysteine residues are involved in forming the regulatory disulfide bridge [Jacquot, J.-P. et al., FEBS Lett. 401 (1997) 143^147] which is the presupposition for the thioredoxin-dependent control of chloroplast FbPase activity. Here we report that each subunit of the FbPase contains an additional structural disulfide bridge which has been observed by combined application of thioredoxins and sulfitolysis. Observation of the structural disulfide bridges by sulfitolysis was only possible when the FbPase was already specifically reduced by the homologous thioredoxin species TRm and TRf from spinach chloroplasts. Interestingly, the accessibility of the structural disulfide bridge for sulfite ions depends on the thioredoxin species engaged in the thioredoxin/FbPase complex. z 1998 Federation of European Biochemical Societies. Key words: Thioredoxin; Sul¢tolysis; Fructose-1,6-bisphosphatase; Protein-protein complex; Disul¢de bridge; Spinacia oleracea
1. Introduction It is well established that thioredoxins are versatile redox proteins involved in the regulation of many di¡erent processes in all kinds of living cells. The redox activity of thioredoxin is due to reversible disul¢de/dithiol interconversions of the active site formed by the tetrapeptide Cys-Gly-Pro-Cys. This enables thioredoxins to receive or to transfer reducing equivalents in biochemical reactions [1]. Higher plants possess multiple thioredoxins which are localized in either cytosol, mitochondria or chloroplasts. Chloroplasts of higher plants usually contain two di¡erent thioredoxin species which are essential for the light-dependent regulation of numerous enzymes. Among them are enzymes of the Calvin-Benson cycle (fructose-1,6-bisphosphatase (FbPase), sedoheptulose-1,7-bisphosphatase, phosphoribulokinase and glyceraldehyde-3phosphate dehydrogenase), sulfate assimilation (adenylylsulfate kinase and 3P-phosphoadenylylsulfate reductase) [2], nitrate assimilation (glutamate synthase) [3], ATP synthase (CF1), glucose-6-phosphate dehydrogenase and NADP-ma*Corresponding author. Fax: (49) (561) 804 4466. E-mail: [email protected] Abbreviations: FbPase, fructose-1,6-bisphosphatase; mBB, monobromobimane; TRf and TRm , thioredoxin species from spinach chloroplasts; DTT, dithiothreitol
late dehydrogenase (NADP-MDH) [2]. The thioredoxin-dependent regulation of all these enzymes has in common that certain regulatory disul¢de bridges are reductively cleaved in a dithiol-disul¢de exchange reaction. The disul¢de bridge of a target enzyme is initially subjected to nucleophilic attack by one of the thiol groups of reduced thioredoxin, resulting in a transient mixed disul¢de bridge between thioredoxin and target enzyme. The adjacent, second thiol group of the active site of thioredoxin cleaves the mixed disul¢de, in generating reduced target enzyme and oxidized thioredoxin [1]. This process induces a conformational change of the enzyme which is accompanied by an activity change. Originally, it was assumed that the oxidized thioredoxin is immediately liberated to be regenerated by the ferredoxin thioredoxin reductase. However, ample evidence is available which indicates that the thioredoxins remain complexed with the target enzyme after the dithiol/disul¢de exchange reaction [4]. The individual proteinprotein interactions between thioredoxin and enzyme are responsible for the ¢ne regulation of the target enzyme activity [5,6]. Moreover, the thioredoxin/enzyme complex protects the enzyme against oxidative inhibition [7]. The redox-active cysteine residues have been identi¢ed in some but not all thioredoxin-dependent enzymes [2]. For example, site-directed mutagenesis studies with pea FbPase indicated that cysteine residues Cys173 and Cys178 form the regulatory disul¢de bridge in each subunit of the tetrameric enzyme [8]. However, the recently published crystal structure of spinach FbPase revealed that the sulfur atoms of these two cysteine residues probably do not form a disul¢de bridge because the distance between the sulphur atoms (Cys174 and Cys179 in spinach) is to large. Most recently Jacquot et al. [10] observed that besides Cys173 and Cys178 cysteine residue Cys153 is also required for redox regulation of pea chloroplast FbPase. This observation raised the question whether in addition to the regulatory disul¢de bridge a structural disul¢de bridge is present per FbPase subunit. Here we present evidence for the existence of such a structural disul¢de bridge in spinach FbPase derived by combination of thioredoxin-dependent reduction and sul¢tolysis. 2. Materials and methods All chemicals and reagents were of highest purity available. The chloroplast thioredoxin species TRm and TRf were generous gifts of Dr. Peter Schuërmann, University of Neuchaêtel, Switzerland. Puri¢cation of spinach chloroplast FbPase followed the procedure by Rother et al. [11]. The incubation mixture for FbPase activation contained in a volume of 0.5 ml, 0.1 M Tris-HCl bu¡er (pH 7.9); 5 mM MgSO4 ; 4 Wg FbPase. Incubation of the FbPase occurred in presence of 1 mM DTT plus thioredoxin and di¡erent concentrations of Na2 SO3 as speci¢ed below. Reductive activation of the enzyme was completed
0014-5793/98/$19.00 ß 1998 Federation of European Biochemical Societies. All rights reserved. PII S 0 0 1 4 - 5 7 9 3 ( 9 8 ) 0 0 1 5 0 - 1
FEBS 19917 5-3-98
110
D.F. Drescher et al./FEBS Letters 424 (1998) 109^112
within 10 min at 30³C. Incubation of the FbPase with sul¢te ions was performed at 30³C for 30 min. Enzyme activity was measured by the addition of 0.05 ml of 60 mM fructose-1,6-bisphosphate solution. After 10 min the reaction was stopped with 0.05 ml 72% trichloroacetic acid, 0.2 ml of the assay mixtures were mixed with 0.8 ml of phosphate reagent, and after 10 min the amount of liberated phosphate was determined photometrically at 660 nm [12]. Fluorescence labeling of liberated thiol groups after reduction and/ or sul¢tolysis (sul¢te incubation) of the FbPase was performed with monobromobimane (mBB from Calbiochem, Thiolyte MQ). 13 Wg of spinach FbPase was incubated in a total volume of 0.3 ml with di¡erent combinations of TRf (18 Wg, 1.4 nmol), TRm (16 Wg, 1.4 nmol), 1 mM DTT and 1 mM Na2 SO3 as speci¢ed below. Reductive incubation was performed at 30³C for 20 min and sul¢tolysis at 30³C for 30 min. The ¢nal labeling step (60 min; 30³C) was initiated by the addition of 0.05 ml of 60 mM mBB solution, and the labeling reaction was stopped with 0.03 ml of 72% trichloroacetic acid. SDS-polyacrylamide gel electrophoresis (10%) was applied to detect mBB labeling of the FbPase which was observed on a £uorescence screen with an excitation wavelength of 366 nm.
3. Results Thioredoxins are speci¢c redox regulators capable of di¡erentiating between regulatory and structural disul¢de bridges in target enzymes [1]. The speci¢city is due to protein-protein interactions in the thioredoxin-enzyme complex [5,6]. Numerous studies have shown that the regulatory disul¢de bridges of the enzymes can be chemically cleaved with low molecular weight reductants such as dithiothreitol. It is obvious that these reductants are not as speci¢c as the physiological thioredoxin reductant which is a protein of about 120 amino acids. Thus, it cannot be excluded that beside regulatory di-
sul¢de bridges additional structural disul¢des will be cleaved by unphysiological chemical reductants. The FbPase from chloroplasts is a tetrameric enzyme, in which each subunit possesses one regulatory disul¢de bridge [9,10]. We established an experimental system which combines the speci¢c reduction of FbPase with the homologous thioredoxins TRm and TRf from spinach chloroplasts with non-enzymatic cleavage of disul¢de bridges by sul¢te ions in order to investigate whether the FbPase contains additional structural disul¢de bridges. Sul¢tolytic cleavage of disul¢de bridges into an Ssulfonyl group and an adjacent thiol group is due to nucleophilic attack by the sul¢te ions. To observe the reduction and/ or the sul¢tolysis of disul¢de bridges the liberated thiol groups were alkylated with mBB, and the incorporation of mBB was followed by its £uorescence emission. We con¢rmed an earlier observation of Droux et al. that oxidized spinach FbPase does not possess an accessible thiol group [15], since labeling of the FbPase with mBB did not occur in the oxidized state (Fig. 1, lanes 1 and 3). In contrast, oxidized TRf is labeled because it contains a third cysteine residue beside the cysteine residues in the active site (Fig. 1, lanes 2 and 3). Addition of 1 mM sodium sul¢te to the oxidized FbPase and TRf caused sul¢tolysis of the disul¢de bridge in the active site of TRf , indicated by the increased mBB labeling (Fig. 1, lanes 5 and 6), but sul¢tolysis of the oxidized FbPase was not observed (Fig. 1, lanes 4 and 6). To observe the thioredoxin-dependent liberation of thiol groups in FbPase the speci¢c thioredoxin reductase was replaced by 1 mM DTT. This DTT concentration was su¤cient for complete reduction of the thioredoxin, but too small to signi¢cantly reduce FbPase (Fig. 1, lanes 8 and
Fig. 1. Liberation of thiol groups in 18 Wg TRf and 13 Wg FbPase from spinach chloroplasts, observed in response to reduction with 1 mM DTT and/or sul¢tolysis with 1 mM sul¢te ions by labeling thiol groups with monobromobimane (mBB). The upper and lower signals correspond to FbPase and TRf , respectively. The reaction conditions are described in Section 2, and the di¡erent reaction mixtures are summarized as follows:
FEBS 19917 5-3-98
D.F. Drescher et al./FEBS Letters 424 (1998) 109^112
111
Fig. 2. Liberation of thiol groups in 16 Wg TRm and 13 Wg FbPase from spinach chloroplasts, observed in response to reduction with 1 mM DTT and/or to sul¢tolysis with 1 mM sul¢te ions by labeling thiol groups with monobromobimane (mBB). The upper and lower signals correspond to FbPase and TRm , respectively. For further details see Fig. 1. The reaction mixtures contained FbPase, TRm and DTT (lane 1) and FbPase, TRm , DTT and sul¢te (lane 2).
9). Labeling of the FbPase by mBB occurred when the enzyme was reduced in presence of 1 mM DTT and 18 Wg TRf (Fig. 1, lane 12). Addition of 1 mM sodium sul¢te to this mixture
produced a signi¢cant increase of the FbPase labeling with mBB (Fig. 1, lane 7). A control experiment evidenced that this labeling was not due to the combined action of DTT and sul¢te ions (Fig. 1, lane 10). In our previous study of TRf saturation kinetics it had been measured that 18 Wg TRf (1.4 nmol) saturate the activation process of FbPase [5]. Thus, the additional liberation of thiol groups in the FbPase by sul¢tolysis, which is not observed with FbPase in the oxidized state, indicates that the cleavage of the regulatory disul¢de bridge by thioredoxin induces a conformational change which makes a buried structural disul¢de susceptible for nucleophilic attack by sul¢te ions. In another set of experiments we studied whether interaction between TRm , the second thioredoxin in spinach chloroplasts, and FbPase produce a comparable conformational change in response to reduction. After treatment with 1 mM DTT and 16 Wg TRm (1.4 nmol) the FbPase was labeled by mBB (Fig. 2, lane 1). In presence of 1 mM sul¢te ions a much smaller increase of the FbPase labeling was observed (Fig. 2; lanes 1 and 2) than in the FbPase/TRf combination (Fig. 1, lanes 7 and 12). Although TRf and TRm possess identical
Fig. 3. In£uence of increasing concentrations of sul¢te ions on the fructose-1,6-bisphosphatase (4 Wg) from spinach chloroplasts before and after speci¢c reduction of the regulatory disul¢de bridges with either TRf (15 Wg) or TRm (15 Wg). The sequence of incubation steps is indicated by the headline of each graph. a, b: The FbPase was incubated together with thioredoxin/DTT and sul¢te ions before the catalytic reaction was initiated by the addition of FbP. c, d: The FbPase was ¢rst treated with sul¢te ions and then incubated with thioredoxin/DTT before the catalytic reaction was initiated by the addition of FbP. E660 of 0.3 is equivalent to a speci¢c FbPase activity of 27.5 Wmol Pi liberated/min/mg protein. The FbPase activity measured without the presence of sul¢te is given at zero mM sul¢te.
FEBS 19917 5-3-98
112
D.F. Drescher et al./FEBS Letters 424 (1998) 109^112
active sites they are not equivalent because of major di¡erences in the amino acid sequences. Obviously, the individual protein-protein interactions between FbPase and TRm or TRf produce di¡erent conformational states of the FbPase. When TRm is bound to FbPase either the structural disul¢de is less accessible for sul¢te ions or the liberated thiol group, which must be labeled with mBB to observe sul¢tolysis, is more buried in this protein complex. The addition of sul¢te ions to FbPase assays containing 1 mM DTT and either TRm or TRf is not inhibitory for FbPase activity (Fig. 3a,b). Moreover, increasing concentrations of sul¢te ions induced a small activity increase and it made no di¡erence whether the sul¢te ions were present during reductive activation or whether incubation with sul¢te ions preceded reductive activation with TRm or TRf (Fig. 3c,d). Thus, sul¢tolysis of the structural disul¢de bridge of the FbPase is not detrimental to its catalytic activity. 4. Discussion Fructose-1,6-bisphosphatase is a key enzyme of the CalvinBenson cycle. The nuclear-encoded enzyme is composed of four identical subunits. The amino acid sequences of the chloroplast FbPases from spinach [16], pea [17], Brassica napus [18], wheat [19], Arabidopsis thaliana [20] and potato [21] have been determined. A signi¢cant extra feature of the chloroplast FbPases is an insertion of about 17 amino acid residues in the middle of the protein containing additional three cysteine residues. Site-directed mutagenesis experiments have revealed that the cysteine residues Cys153 , Cys173 , and Cys178 (in pea FbPase numbering) are involved in the thioredoxindependent redox regulation, but it was not clari¢ed which pairing of these three cysteine residues forms the regulatory disul¢de bridge [10]. The three-dimensional crystal structure of the FbPase from spinach chloroplasts shows that the inserted peptide is localized at the outside corner of each subunit [9]. This exposed position confers the redox regulation to chloroplast FbPase, and mutagenesis experiments have shown that the docking site for thioredoxin precedes the regulatory sequence [22]. It was recently suggested that an additional structural disul¢de bridge should be present in each subunit of the FbPase [10]. We applied a combination of speci¢c opening of the regulatory disul¢de bridge by thioredoxins and chemical cleavage of other disul¢de bridges by sul¢te ions. The latter reaction, which is called sul¢tolysis, has been recognized to be responsible for the phytotoxic action of SO2 in higher plants [13,14]. Liberated thiol groups were labeled with monobromobimane, and incorporation of the label was observed by the £uorescence emission of proteinbound mBB. The results demonstrate that oxidized FbPase possesses a buried structural disul¢de bridge which becomes accessible for sul¢tolysis only when the regulatory disul¢de bridge was speci¢cally reduced by thioredoxin (Figs. 1 and 2). Cleavage of this structural disul¢de bridge by sul¢te ions did not signi¢cantly in£uence the catalytic activity of the FbPase (Fig. 3) which indicates that this disul¢de bridge is not involved in the activation process. Interestingly, the accessibility of the structural disul¢de bridge was dependent on the thioredoxin species applied in the experiments. Sul¢tolysis occurred to a much greater extent in combination with TRf than with TRm (Figs. 1 and 2). Obviously, the protein-protein interactions between FbPase and the individual thioredoxins
cause di¡erent FbPase conformations. This is further evidence that the formation of a thioredoxin-target enzyme complex is responsible for the ¢ne regulation of enzyme activity [1]. Moreover, the results of protein labeling, e.g. the detection of disul¢de bridges, depend on the individual applied thioredoxin species. Thus, only homologous systems comprising thioredoxins and target enzymes from the same plant and cell organelle will completely describe the regulatory action of thioredoxins. The data presented here indicate that the conformation of the FbPase is £exible and that crystal packing forces have probably generated a three-dimensional structure in which the distance between the cysteine residues is too long for a regulatory bridge in spinach chloroplast FbPase [9]. Our ¢nding of a structural disul¢de bridge besides the regulatory one complicates the identi¢cation of the individual cysteine residues which are involved in the regulatory, or structural disul¢de bridges, respectively. It was recently suggested that cysteine residues Cys49 and Cys190 (pea FbPase) may form a disul¢de bridge [10], but further experiments are necessary to evaluate this suggestion. Acknowledgements: Support from the Deutsche Forschungsgemeinschaft (I.H.) is gratefully acknowledged.
References [1] Follmann, H. and Haëberlein, I. (1996) BioFactors 5, 147^156. [2] Jacquot, J.P., Lancelin, J.-M. and Meyer, Y. (1997) New Phytol. 136, 543^570. [3] Lichter, A. and Haëberlein, I. (1998) J. Plant Physiol. (in press). [4] Haëberlein, I., Wuërfel, M. and Follmann, H. (1992) Biochim. Biophys. Acta 1121, 293^296. [5] Haëberlein, I. and Vogeler, B. (1995) Biochim. Biophys. Acta 1253, 169^174. [6] Braun, H., Lichter, A. and Haëberlein, I. (1996) Eur. J. Biochem. 240, 781^788. [7] Morell, S., Follmann, H., De Tullio, M. and Haëberlein, I. (1997) FEBS Lett. 414, 567^579. [8] Jacquot, J.P., Lopez-Jaramillo, J., Chueca, A., Cher¢ls, J., Lemaire, S., Chedozeau, B., Miginiac-Maslow, M., Decottingnies, P., Wolosiuk, R.A. and Lopez-Gorge, J. (1995) Eur. J. Biochem. 229, 675^681. [9] Villeret, V., Huang, S., Zhang, Y., Xue, Y. and Lipscomb, W.N. (1995) Biochemistry 34, 4299^4306. [10] Jacquot, J.P., Lopez-Jaramillo, J., Miginiac-Maslow, M., Lemaire, S., Cher¢ls, J., Chueca, A. and Lopez-Gorge, J. (1997) FEBS Lett. 401, 143^147. [11] Rother, Th., Acker, G. and Scheibe, R. (1988) Bot. Acta 101, 311^320. [12] Taussky, H. and Shorr, H.E. (1963) J. Biol. Chem. 202, 675^685. [13] Haëberlein, I. (1994) Eur. J. Biochem. 223, 473^479. [14] Haëberlein, I. and Follmann, H. (1991) Naturwissenschaften 78, 259^265. [15] Droux, M., Crawford, N.A. and Buchanan, B.B. (1987) C.R. Acad. Sci. Paris Seèr. III Physiol. Veèg. 305, 335^341. [16] Marcus, F., Edelstein, I., Reardon, I. and Heinrikson, R.L. (1982) Proc. Natl. Acad. Sci. USA 79, 7161^7165. [17] Carrasco, J.L., Chueca, A., Prado, F.E., Hermoso, R., Lazaro, J.J., Ramos, J.L., Sahrawy, M. and Lopez-Gorge, J. (1994) Planta 193, 494^501. [18] Rodriquez-Suarez, R.J. and Wolosiuk, R.A. (1993) Plant Physiol. 103, 1453^1454. [19] Raines, C.A., Lloyd, J.C., Willingham, N.M., Potts, S. and Dyer, T.A. (1992) Eur. J. Biochem. 205, 1053^1059. [20] Horsnell, P.R. and Raines, C.A. (1991) Plant Mol. Biol. 17, 185^ 186. [21] Kossmann, J., Muëller-Roëber, B., Dyer, T.A., Raines, C.A., Sonnewald, U. and Willmitzer, L. (1992) Planta 188, 7^12. [22] Sahrawy, M., Chueca, A., Hermoso, R., Lazaro, J.J. and LopezGorge, J. (1997) J. Mol. Biol. 269, 623^630.
FEBS 19917 5-3-98
FEBS Letters 424 (1998) 109^112
Sul¢tolysis and thioredoxin-dependent reduction reveal the presence of a structural disul¢de bridge in spinach chloroplast fructose-1,6-bisphosphatase Dorothee F. Drescher, Hartmut Follmann, Ingo Haëberlein* University of Kassel, Department of Biochemistry, Heinrich-Plett-Str. 40, D-34109 Kassel, Germany Received 28 November 1997; revised version received 30 January 1998
Abstract A significant difference between cytosolic and chloroplastic fructose-1,6-bisphosphatase (FbPase) is an extra peptide in the middle of chloroplast FbPase which contains three additional cysteine residues. Sit-directed mutagenesis experiments have shown that at least two of these cysteine residues are involved in forming the regulatory disulfide bridge [Jacquot, J.-P. et al., FEBS Lett. 401 (1997) 143^147] which is the presupposition for the thioredoxin-dependent control of chloroplast FbPase activity. Here we report that each subunit of the FbPase contains an additional structural disulfide bridge which has been observed by combined application of thioredoxins and sulfitolysis. Observation of the structural disulfide bridges by sulfitolysis was only possible when the FbPase was already specifically reduced by the homologous thioredoxin species TRm and TRf from spinach chloroplasts. Interestingly, the accessibility of the structural disulfide bridge for sulfite ions depends on the thioredoxin species engaged in the thioredoxin/FbPase complex. z 1998 Federation of European Biochemical Societies. Key words: Thioredoxin; Sul¢tolysis; Fructose-1,6-bisphosphatase; Protein-protein complex; Disul¢de bridge; Spinacia oleracea
1. Introduction It is well established that thioredoxins are versatile redox proteins involved in the regulation of many di¡erent processes in all kinds of living cells. The redox activity of thioredoxin is due to reversible disul¢de/dithiol interconversions of the active site formed by the tetrapeptide Cys-Gly-Pro-Cys. This enables thioredoxins to receive or to transfer reducing equivalents in biochemical reactions [1]. Higher plants possess multiple thioredoxins which are localized in either cytosol, mitochondria or chloroplasts. Chloroplasts of higher plants usually contain two di¡erent thioredoxin species which are essential for the light-dependent regulation of numerous enzymes. Among them are enzymes of the Calvin-Benson cycle (fructose-1,6-bisphosphatase (FbPase), sedoheptulose-1,7-bisphosphatase, phosphoribulokinase and glyceraldehyde-3phosphate dehydrogenase), sulfate assimilation (adenylylsulfate kinase and 3P-phosphoadenylylsulfate reductase) [2], nitrate assimilation (glutamate synthase) [3], ATP synthase (CF1), glucose-6-phosphate dehydrogenase and NADP-ma*Corresponding author. Fax: (49) (561) 804 4466. E-mail: [email protected] Abbreviations: FbPase, fructose-1,6-bisphosphatase; mBB, monobromobimane; TRf and TRm , thioredoxin species from spinach chloroplasts; DTT, dithiothreitol
late dehydrogenase (NADP-MDH) [2]. The thioredoxin-dependent regulation of all these enzymes has in common that certain regulatory disul¢de bridges are reductively cleaved in a dithiol-disul¢de exchange reaction. The disul¢de bridge of a target enzyme is initially subjected to nucleophilic attack by one of the thiol groups of reduced thioredoxin, resulting in a transient mixed disul¢de bridge between thioredoxin and target enzyme. The adjacent, second thiol group of the active site of thioredoxin cleaves the mixed disul¢de, in generating reduced target enzyme and oxidized thioredoxin [1]. This process induces a conformational change of the enzyme which is accompanied by an activity change. Originally, it was assumed that the oxidized thioredoxin is immediately liberated to be regenerated by the ferredoxin thioredoxin reductase. However, ample evidence is available which indicates that the thioredoxins remain complexed with the target enzyme after the dithiol/disul¢de exchange reaction [4]. The individual proteinprotein interactions between thioredoxin and enzyme are responsible for the ¢ne regulation of the target enzyme activity [5,6]. Moreover, the thioredoxin/enzyme complex protects the enzyme against oxidative inhibition [7]. The redox-active cysteine residues have been identi¢ed in some but not all thioredoxin-dependent enzymes [2]. For example, site-directed mutagenesis studies with pea FbPase indicated that cysteine residues Cys173 and Cys178 form the regulatory disul¢de bridge in each subunit of the tetrameric enzyme [8]. However, the recently published crystal structure of spinach FbPase revealed that the sulfur atoms of these two cysteine residues probably do not form a disul¢de bridge because the distance between the sulphur atoms (Cys174 and Cys179 in spinach) is to large. Most recently Jacquot et al. [10] observed that besides Cys173 and Cys178 cysteine residue Cys153 is also required for redox regulation of pea chloroplast FbPase. This observation raised the question whether in addition to the regulatory disul¢de bridge a structural disul¢de bridge is present per FbPase subunit. Here we present evidence for the existence of such a structural disul¢de bridge in spinach FbPase derived by combination of thioredoxin-dependent reduction and sul¢tolysis. 2. Materials and methods All chemicals and reagents were of highest purity available. The chloroplast thioredoxin species TRm and TRf were generous gifts of Dr. Peter Schuërmann, University of Neuchaêtel, Switzerland. Puri¢cation of spinach chloroplast FbPase followed the procedure by Rother et al. [11]. The incubation mixture for FbPase activation contained in a volume of 0.5 ml, 0.1 M Tris-HCl bu¡er (pH 7.9); 5 mM MgSO4 ; 4 Wg FbPase. Incubation of the FbPase occurred in presence of 1 mM DTT plus thioredoxin and di¡erent concentrations of Na2 SO3 as speci¢ed below. Reductive activation of the enzyme was completed
0014-5793/98/$19.00 ß 1998 Federation of European Biochemical Societies. All rights reserved. PII S 0 0 1 4 - 5 7 9 3 ( 9 8 ) 0 0 1 5 0 - 1
FEBS 19917 5-3-98
110
D.F. Drescher et al./FEBS Letters 424 (1998) 109^112
within 10 min at 30³C. Incubation of the FbPase with sul¢te ions was performed at 30³C for 30 min. Enzyme activity was measured by the addition of 0.05 ml of 60 mM fructose-1,6-bisphosphate solution. After 10 min the reaction was stopped with 0.05 ml 72% trichloroacetic acid, 0.2 ml of the assay mixtures were mixed with 0.8 ml of phosphate reagent, and after 10 min the amount of liberated phosphate was determined photometrically at 660 nm [12]. Fluorescence labeling of liberated thiol groups after reduction and/ or sul¢tolysis (sul¢te incubation) of the FbPase was performed with monobromobimane (mBB from Calbiochem, Thiolyte MQ). 13 Wg of spinach FbPase was incubated in a total volume of 0.3 ml with di¡erent combinations of TRf (18 Wg, 1.4 nmol), TRm (16 Wg, 1.4 nmol), 1 mM DTT and 1 mM Na2 SO3 as speci¢ed below. Reductive incubation was performed at 30³C for 20 min and sul¢tolysis at 30³C for 30 min. The ¢nal labeling step (60 min; 30³C) was initiated by the addition of 0.05 ml of 60 mM mBB solution, and the labeling reaction was stopped with 0.03 ml of 72% trichloroacetic acid. SDS-polyacrylamide gel electrophoresis (10%) was applied to detect mBB labeling of the FbPase which was observed on a £uorescence screen with an excitation wavelength of 366 nm.
3. Results Thioredoxins are speci¢c redox regulators capable of di¡erentiating between regulatory and structural disul¢de bridges in target enzymes [1]. The speci¢city is due to protein-protein interactions in the thioredoxin-enzyme complex [5,6]. Numerous studies have shown that the regulatory disul¢de bridges of the enzymes can be chemically cleaved with low molecular weight reductants such as dithiothreitol. It is obvious that these reductants are not as speci¢c as the physiological thioredoxin reductant which is a protein of about 120 amino acids. Thus, it cannot be excluded that beside regulatory di-
sul¢de bridges additional structural disul¢des will be cleaved by unphysiological chemical reductants. The FbPase from chloroplasts is a tetrameric enzyme, in which each subunit possesses one regulatory disul¢de bridge [9,10]. We established an experimental system which combines the speci¢c reduction of FbPase with the homologous thioredoxins TRm and TRf from spinach chloroplasts with non-enzymatic cleavage of disul¢de bridges by sul¢te ions in order to investigate whether the FbPase contains additional structural disul¢de bridges. Sul¢tolytic cleavage of disul¢de bridges into an Ssulfonyl group and an adjacent thiol group is due to nucleophilic attack by the sul¢te ions. To observe the reduction and/ or the sul¢tolysis of disul¢de bridges the liberated thiol groups were alkylated with mBB, and the incorporation of mBB was followed by its £uorescence emission. We con¢rmed an earlier observation of Droux et al. that oxidized spinach FbPase does not possess an accessible thiol group [15], since labeling of the FbPase with mBB did not occur in the oxidized state (Fig. 1, lanes 1 and 3). In contrast, oxidized TRf is labeled because it contains a third cysteine residue beside the cysteine residues in the active site (Fig. 1, lanes 2 and 3). Addition of 1 mM sodium sul¢te to the oxidized FbPase and TRf caused sul¢tolysis of the disul¢de bridge in the active site of TRf , indicated by the increased mBB labeling (Fig. 1, lanes 5 and 6), but sul¢tolysis of the oxidized FbPase was not observed (Fig. 1, lanes 4 and 6). To observe the thioredoxin-dependent liberation of thiol groups in FbPase the speci¢c thioredoxin reductase was replaced by 1 mM DTT. This DTT concentration was su¤cient for complete reduction of the thioredoxin, but too small to signi¢cantly reduce FbPase (Fig. 1, lanes 8 and
Fig. 1. Liberation of thiol groups in 18 Wg TRf and 13 Wg FbPase from spinach chloroplasts, observed in response to reduction with 1 mM DTT and/or sul¢tolysis with 1 mM sul¢te ions by labeling thiol groups with monobromobimane (mBB). The upper and lower signals correspond to FbPase and TRf , respectively. The reaction conditions are described in Section 2, and the di¡erent reaction mixtures are summarized as follows:
FEBS 19917 5-3-98
D.F. Drescher et al./FEBS Letters 424 (1998) 109^112
111
Fig. 2. Liberation of thiol groups in 16 Wg TRm and 13 Wg FbPase from spinach chloroplasts, observed in response to reduction with 1 mM DTT and/or to sul¢tolysis with 1 mM sul¢te ions by labeling thiol groups with monobromobimane (mBB). The upper and lower signals correspond to FbPase and TRm , respectively. For further details see Fig. 1. The reaction mixtures contained FbPase, TRm and DTT (lane 1) and FbPase, TRm , DTT and sul¢te (lane 2).
9). Labeling of the FbPase by mBB occurred when the enzyme was reduced in presence of 1 mM DTT and 18 Wg TRf (Fig. 1, lane 12). Addition of 1 mM sodium sul¢te to this mixture
produced a signi¢cant increase of the FbPase labeling with mBB (Fig. 1, lane 7). A control experiment evidenced that this labeling was not due to the combined action of DTT and sul¢te ions (Fig. 1, lane 10). In our previous study of TRf saturation kinetics it had been measured that 18 Wg TRf (1.4 nmol) saturate the activation process of FbPase [5]. Thus, the additional liberation of thiol groups in the FbPase by sul¢tolysis, which is not observed with FbPase in the oxidized state, indicates that the cleavage of the regulatory disul¢de bridge by thioredoxin induces a conformational change which makes a buried structural disul¢de susceptible for nucleophilic attack by sul¢te ions. In another set of experiments we studied whether interaction between TRm , the second thioredoxin in spinach chloroplasts, and FbPase produce a comparable conformational change in response to reduction. After treatment with 1 mM DTT and 16 Wg TRm (1.4 nmol) the FbPase was labeled by mBB (Fig. 2, lane 1). In presence of 1 mM sul¢te ions a much smaller increase of the FbPase labeling was observed (Fig. 2; lanes 1 and 2) than in the FbPase/TRf combination (Fig. 1, lanes 7 and 12). Although TRf and TRm possess identical
Fig. 3. In£uence of increasing concentrations of sul¢te ions on the fructose-1,6-bisphosphatase (4 Wg) from spinach chloroplasts before and after speci¢c reduction of the regulatory disul¢de bridges with either TRf (15 Wg) or TRm (15 Wg). The sequence of incubation steps is indicated by the headline of each graph. a, b: The FbPase was incubated together with thioredoxin/DTT and sul¢te ions before the catalytic reaction was initiated by the addition of FbP. c, d: The FbPase was ¢rst treated with sul¢te ions and then incubated with thioredoxin/DTT before the catalytic reaction was initiated by the addition of FbP. E660 of 0.3 is equivalent to a speci¢c FbPase activity of 27.5 Wmol Pi liberated/min/mg protein. The FbPase activity measured without the presence of sul¢te is given at zero mM sul¢te.
FEBS 19917 5-3-98
112
D.F. Drescher et al./FEBS Letters 424 (1998) 109^112
active sites they are not equivalent because of major di¡erences in the amino acid sequences. Obviously, the individual protein-protein interactions between FbPase and TRm or TRf produce di¡erent conformational states of the FbPase. When TRm is bound to FbPase either the structural disul¢de is less accessible for sul¢te ions or the liberated thiol group, which must be labeled with mBB to observe sul¢tolysis, is more buried in this protein complex. The addition of sul¢te ions to FbPase assays containing 1 mM DTT and either TRm or TRf is not inhibitory for FbPase activity (Fig. 3a,b). Moreover, increasing concentrations of sul¢te ions induced a small activity increase and it made no di¡erence whether the sul¢te ions were present during reductive activation or whether incubation with sul¢te ions preceded reductive activation with TRm or TRf (Fig. 3c,d). Thus, sul¢tolysis of the structural disul¢de bridge of the FbPase is not detrimental to its catalytic activity. 4. Discussion Fructose-1,6-bisphosphatase is a key enzyme of the CalvinBenson cycle. The nuclear-encoded enzyme is composed of four identical subunits. The amino acid sequences of the chloroplast FbPases from spinach [16], pea [17], Brassica napus [18], wheat [19], Arabidopsis thaliana [20] and potato [21] have been determined. A signi¢cant extra feature of the chloroplast FbPases is an insertion of about 17 amino acid residues in the middle of the protein containing additional three cysteine residues. Site-directed mutagenesis experiments have revealed that the cysteine residues Cys153 , Cys173 , and Cys178 (in pea FbPase numbering) are involved in the thioredoxindependent redox regulation, but it was not clari¢ed which pairing of these three cysteine residues forms the regulatory disul¢de bridge [10]. The three-dimensional crystal structure of the FbPase from spinach chloroplasts shows that the inserted peptide is localized at the outside corner of each subunit [9]. This exposed position confers the redox regulation to chloroplast FbPase, and mutagenesis experiments have shown that the docking site for thioredoxin precedes the regulatory sequence [22]. It was recently suggested that an additional structural disul¢de bridge should be present in each subunit of the FbPase [10]. We applied a combination of speci¢c opening of the regulatory disul¢de bridge by thioredoxins and chemical cleavage of other disul¢de bridges by sul¢te ions. The latter reaction, which is called sul¢tolysis, has been recognized to be responsible for the phytotoxic action of SO2 in higher plants [13,14]. Liberated thiol groups were labeled with monobromobimane, and incorporation of the label was observed by the £uorescence emission of proteinbound mBB. The results demonstrate that oxidized FbPase possesses a buried structural disul¢de bridge which becomes accessible for sul¢tolysis only when the regulatory disul¢de bridge was speci¢cally reduced by thioredoxin (Figs. 1 and 2). Cleavage of this structural disul¢de bridge by sul¢te ions did not signi¢cantly in£uence the catalytic activity of the FbPase (Fig. 3) which indicates that this disul¢de bridge is not involved in the activation process. Interestingly, the accessibility of the structural disul¢de bridge was dependent on the thioredoxin species applied in the experiments. Sul¢tolysis occurred to a much greater extent in combination with TRf than with TRm (Figs. 1 and 2). Obviously, the protein-protein interactions between FbPase and the individual thioredoxins
cause di¡erent FbPase conformations. This is further evidence that the formation of a thioredoxin-target enzyme complex is responsible for the ¢ne regulation of enzyme activity [1]. Moreover, the results of protein labeling, e.g. the detection of disul¢de bridges, depend on the individual applied thioredoxin species. Thus, only homologous systems comprising thioredoxins and target enzymes from the same plant and cell organelle will completely describe the regulatory action of thioredoxins. The data presented here indicate that the conformation of the FbPase is £exible and that crystal packing forces have probably generated a three-dimensional structure in which the distance between the cysteine residues is too long for a regulatory bridge in spinach chloroplast FbPase [9]. Our ¢nding of a structural disul¢de bridge besides the regulatory one complicates the identi¢cation of the individual cysteine residues which are involved in the regulatory, or structural disul¢de bridges, respectively. It was recently suggested that cysteine residues Cys49 and Cys190 (pea FbPase) may form a disul¢de bridge [10], but further experiments are necessary to evaluate this suggestion. Acknowledgements: Support from the Deutsche Forschungsgemeinschaft (I.H.) is gratefully acknowledged.
References [1] Follmann, H. and Haëberlein, I. (1996) BioFactors 5, 147^156. [2] Jacquot, J.P., Lancelin, J.-M. and Meyer, Y. (1997) New Phytol. 136, 543^570. [3] Lichter, A. and Haëberlein, I. (1998) J. Plant Physiol. (in press). [4] Haëberlein, I., Wuërfel, M. and Follmann, H. (1992) Biochim. Biophys. Acta 1121, 293^296. [5] Haëberlein, I. and Vogeler, B. (1995) Biochim. Biophys. Acta 1253, 169^174. [6] Braun, H., Lichter, A. and Haëberlein, I. (1996) Eur. J. Biochem. 240, 781^788. [7] Morell, S., Follmann, H., De Tullio, M. and Haëberlein, I. (1997) FEBS Lett. 414, 567^579. [8] Jacquot, J.P., Lopez-Jaramillo, J., Chueca, A., Cher¢ls, J., Lemaire, S., Chedozeau, B., Miginiac-Maslow, M., Decottingnies, P., Wolosiuk, R.A. and Lopez-Gorge, J. (1995) Eur. J. Biochem. 229, 675^681. [9] Villeret, V., Huang, S., Zhang, Y., Xue, Y. and Lipscomb, W.N. (1995) Biochemistry 34, 4299^4306. [10] Jacquot, J.P., Lopez-Jaramillo, J., Miginiac-Maslow, M., Lemaire, S., Cher¢ls, J., Chueca, A. and Lopez-Gorge, J. (1997) FEBS Lett. 401, 143^147. [11] Rother, Th., Acker, G. and Scheibe, R. (1988) Bot. Acta 101, 311^320. [12] Taussky, H. and Shorr, H.E. (1963) J. Biol. Chem. 202, 675^685. [13] Haëberlein, I. (1994) Eur. J. Biochem. 223, 473^479. [14] Haëberlein, I. and Follmann, H. (1991) Naturwissenschaften 78, 259^265. [15] Droux, M., Crawford, N.A. and Buchanan, B.B. (1987) C.R. Acad. Sci. Paris Seèr. III Physiol. Veèg. 305, 335^341. [16] Marcus, F., Edelstein, I., Reardon, I. and Heinrikson, R.L. (1982) Proc. Natl. Acad. Sci. USA 79, 7161^7165. [17] Carrasco, J.L., Chueca, A., Prado, F.E., Hermoso, R., Lazaro, J.J., Ramos, J.L., Sahrawy, M. and Lopez-Gorge, J. (1994) Planta 193, 494^501. [18] Rodriquez-Suarez, R.J. and Wolosiuk, R.A. (1993) Plant Physiol. 103, 1453^1454. [19] Raines, C.A., Lloyd, J.C., Willingham, N.M., Potts, S. and Dyer, T.A. (1992) Eur. J. Biochem. 205, 1053^1059. [20] Horsnell, P.R. and Raines, C.A. (1991) Plant Mol. Biol. 17, 185^ 186. [21] Kossmann, J., Muëller-Roëber, B., Dyer, T.A., Raines, C.A., Sonnewald, U. and Willmitzer, L. (1992) Planta 188, 7^12. [22] Sahrawy, M., Chueca, A., Hermoso, R., Lazaro, J.J. and LopezGorge, J. (1997) J. Mol. Biol. 269, 623^630.
FEBS 19917 5-3-98